Concern 81 · Working draft · editorial review in progress
Cultural bias in annotation guidelines
Cultural bias in annotation guidelines occurs when the rules for labeling data assume one cultural interpretation applies to everyone.
Understanding the concern
Guidelines decide what counts as polite, relevant, harmful, or normal. Those judgments can depend on context. Workers may follow the instructions perfectly while producing labels that fit the guideline writer's community better than the eventual users.
This differs from workers making inconsistent choices. The consistency itself can preserve the narrow rule. Review should involve people who understand the affected settings and make room for disagreement where one interpretation is not enough.
An example
A guideline may label direct speech as rude even where users understand it as ordinary, respectful communication.
Illustrative unless identified as a documented case.
Words in plain English
- Annotation
- Human work that labels or describes examples for an AI system. Labels can involve judgment and can be wrong.
Evidence and its limits
Supports attention to subjective interpretation and whose preferences models reflect; the example is hypothetical.
Barbara Plank · EMNLP 2022 · Research position paper
The Problem of Human Label Variation: On Ground Truth in Data, Modeling and Evaluation ↗ (opens in a new tab)Explains why human disagreement can contain useful information rather than merely label noise.
Santurkar and colleagues · ICML 2023 · Research study
Whose Opinions Do Language Models Reflect? ↗ (opens in a new tab)Compares model responses with US public-opinion groups. It does not isolate training data as the sole cause of differences.
Broader frameworks and background
These resources provide context. They do not establish every claim in this concern.
A question to discuss
Whose cultural assumptions are built into the labeling rules, and who helped review them?
Keep track of different views, what needs checking, and any next step in your own discussion notes.
Open discussion notes →Principles for the conversation
Before you finish · reflect together
Virtues and principles to consider
Virtues are qualities we practice in how we treat people and make choices. They are not scores for judging someone. Use these ideas to explore what a thoughtful response could look like.
Truthfulness
Truthfulness means making claims that match the evidence and being clear about what we do not know. It includes correcting mistakes rather than protecting a polished story.
Check sources, label guesses, and distinguish a trial result from a proven benefit. If AI helped produce a claim, someone still needs to take responsibility for checking it.
Ask together: What can we honestly say, and what still needs checking?
Justice
Justice means fair treatment, fair procedures, and attention to who receives benefits and burdens. It asks whether people with less power can be heard and seek repair.
Look at outcomes across different groups and individual cases. Provide a usable way to challenge decisions and correct harm. Equal wording in a policy does not guarantee equal treatment in practice.
Ask together: Who benefits, who bears the cost, and who can challenge the decision?
Stewardship
Stewardship means caring for resources and systems that others depend on, including people who will use them later. It looks beyond the benefit of a quick launch.
Plan for maintenance, energy use, data protection, and eventual shutdown. Ask who will look after the tool when the original team moves on and how future users can avoid inheriting hidden costs.
Ask together: What are we responsible for caring for over time?
Editorial ratings
These are editorial judgments from the register, not measured probabilities or predictions about your situation.
About the register and scoring →More ways to explore
Search for more writing
For more specific or recent material, open a search below or ask for a web summary. AI-generated summaries can contain mistakes; open the original articles to check them.